|
In many highly congested terminal areas, air
traffic controllers often do not provide optimal
aircraft runway assignment, sequencing, and spacing
while maintaining the public's expected level of
safety. NASA and the Federal Aviation Administration
(FAA) are continuing to design, develop, and deploy
a software-based decision support tool (DST), called
the Final Approach Spacing Tool (FAST), for terminal-area
air traffic management and control of arrival
aircraft. FAST incorporates advanced-knowledge
engineering algorithms, accurate trajectory prediction,
and a specialized graphical-user interface (GUI)
to provide detailed schedule information to TRACON
traffic management coordinators and commands to
TRACON approach controllers. An early version of
this DST, known as Passive FAST (pFAST), provided a
reduced set of advisories, namely runway assign-ments
and landing sequences. These advisories
enabled controllers to achieve a more balanced
airport, as well as throughput increases of 9%-13%.
The follow-on version of this DST, known as Active
FAST (aFAST), will provide additional tactical advisory
information, namely, heading, speed, and
altitude commands. These advisories, if followed,
allow reduction of excess-in-trail separation between
aircraft at touchdown. Continued research is required
to enhance FAST so that it can provide this additional
advisory information. This research includes the
development of advanced scheduling algorithms and
automated conflict-resolution schemes and the
definition of an appropriate computer-human
interface (CHI).
Figure 1 shows the aFAST GUI with a typical set
of passive and active advisories. The passive advisories,
runway assignment and landing sequence, are
presented as yellow text in the aircraft's full
datablock (FDB). In this example, AAL1109 is
sequenced No. 2 to runway 18R. The active advisories,
heading, and speed are presented as cyan and
orange map graphics and text in the aircraft's FDB.
In figure 1, the controller is being advised to start
turning EGF423 to 100 degrees at the filled diamond
and to start slowing DAL1105 to 210 knots IAS at the
unfilled circle.
During FY99, a significant amount of the aFAST
software infrastructure was completed. The new
design addresses several key limitations discovered
during the operational testing and deployment of
pFAST, and allows easier rapid prototyping of
sequencing and conflict resolution logic. In addition,
a method for investigating CHI requirements of the
aFAST GUI, independent of the aFAST scheduling
algorithms, was developed. This approach provided
controllers with active advisory information replayed
from recorded traffic scenarios. Though the controllers
are not actively controlling the aircraft, they issue
the advisories and evaluate the user interface. This
method decouples the evaluations of the user interface
and scheduling algorithms, while maintaining a
realistic air traffic environment. A series of "shadow"
simulations was conducted to evaluate advisory
format, symbols, timing, and use of color. During
these simulations, controller reaction times to
advisory onset and command issuance were
recorded. Following each scenario, questionnaires
were administered to assess the usability of and
workload associated with aFAST advisories. Results
from the study comparing color and monochrome
advisory presentation indicate that controllers noticed
advisory onset more quickly when those advisories
were presented in color. Controllers also rated the
color advisories as producing less screen clutter and
lower mental workload.
Point of Contact: J. Robinson
(650) 604-0873
jerobinson@mail.arc.nasa.gov
Back To Top
Previous Paper
Return to Global Civil Aviation
Next Paper |
|
Fig. 1. aFAST graphical-user interface showing typical advisories.
|
|
